<p>What I wonder–though with no evidence to support the wondering–is whether colleges unconsciously discriminate against “typical” Asian applicants in math/science/engineering/business. That is, would a white (or URM) candidate in those fields with similar qualifications be looked upon more favorably than an Asian candidate? Even if the Asian candidate would be slightly more qualified in a race-blind consideration? Given current stereotypes, this seems plausible.</p>
<p>I would look at the Duke Study from an alternative PoV. [News:</a> Testing for ‘Mismatch’ - Inside Higher Ed](<a href=“http://www.insidehighered.com/news/2009/04/20/mismatch]News:”>http://www.insidehighered.com/news/2009/04/20/mismatch) Instead of comparing Asian students with URMs , let’s compare them with Caucasian students instead. Here are the scores assigned by the Duke adcom.</p>
<pre><code> Caucasian Asian
</code></pre>
<p>–Achievement 4.34 4.67
–Curriculum 4.71 4.91
–Essay 3.52 3.58
–Personal qualities 3.57 3.52
–Recommendations 3.97 4.06
–Test scores 3.69 4.10</p>
<p>Taking into account that the disciplines popular to Asian students in the elites are also more competitive, such as engineering, science, business, and economics, I would say the above scores are close enough that it is reasonable to interpret that Asian students aren’t significantly discriminated against their Caucasian counterparts at Duke, i.e. some discrimination is still possible, but not overwhelming. It is only the URMs who are being favorably handled against all other ethnic groups.</p>
<p>Now if we were to consider Duke as a proxy for an elite with national appeal and without significant discrimination against Asian students, we may invoke “The Duke Test” ----- when Duke can accommodate about 22 % Asian undergraduates (and about 5 % “unknowns”), should other elites be able to do similarly (or better)? Caltech, MIT, and Stanford would pass the test. Harvard with about 17 % Asian undergraduates may still pass the test, if we give them the benefit of a doubt that a good portion of the “unknowns” are in fact Asian students. But Yale and Princeton look iffy, unless about half of the “unknowns” are in fact Asian students.</p>
<p>
Yale is not so much of a math/engineering school. A lot of Asian kids who could get in Yale were usually cross-admitted with other HYPSM, and did not end up with Yale. I could be wrong, but from what I have seen in the past several years, it seems to be the case.</p>
<p>
</p>
<p>And a lot are admitted each year to Princeton, are cross-admits to H, and choose to matriculate to H.</p>
<p>Acceptances =/= Yield.</p>
<p>Keilexandra, I’m pretty busy for the next 24 hours. Will respond later this weekend.</p>
<p>Lol is Yale really that hard for Asians to get in?</p>
<p>Just stock up on leaderships and test scores right?</p>
<p>
Who said that it is hard to get in? Many got in, just did not attend.</p>
<p>
</p>
<p>No. Those who could get in usually get close to perfect SAT scores at 8th grade, and have many things at the state level at the time they apply to colleges. They choose the schools, not the schools choose them.</p>
<p>The people who take this seriously are a sad, sad bunch. Chill out. Stop denying that racial discrimination is not only necessary but also natural and therefore healthy.</p>
<p>It is sad and disheartening that some people actually believe racial discrimination is necessary, natural, and healthy.</p>
<p>Although racial discrimination may be natural, it is not necessary nor is it healthy.</p>
<p>kidding, sort of ;)</p>
<p>
</p>
<p>Give that man a prize. As I mentioned earlier (post #530), the Duke data is actually evidence against that school discriminating between white and Asian applicants. Engineering enrollment is one of a half-dozen factors driving a white/Asian credential disparity (among matriculants) including early decision, legacy preference, athlete preference, and a huge enrollment quota for North Carolina.</p>
<p>That’s just the Duke data. The Duke study concluded that students (and their families, friends, co-ethnics, etc) have no basis for assessing their own abilities, beyond the information they reveal to the admissions officers — but that admissions officers do have a further basis for making additional judgements of student ability (information going beyond that provided in the application). According to the study, the admissions office has a quantifiable form of “private information”, whereas the student’s private information is worthless.</p>
<p>The application of that finding to CC discussions is obvious. If Duke rejects Lian Ji with his 799.99 SAT average and 3.99997 GPA, causing a firestorm on Chinese message boards the world over, the Duke study would simply suggest that the admissions office can (on the average, across the applicant pool) accurately see something about Lian’s abilities, that the Chinese parents and their children are blind to.</p>
<p>^ Can you elaborate on the concept that majors popular among Asian students are more “competitive”? How is one major more competitive than another major, aside from discrepancies in department size? For instance, I can understand that a classics major might be more valued than an English major; but I don’t see that an English major is any more valued than a bio major.</p>
<p>If Asian students are rated higher on both quantitative and qualitative factors than white students, and white students enroll in a greater diversity of majors, and Asians are not being discriminated against, then it follows that Asians should represent almost all of the students enrolled in the “popular Asian majors” because they are overall best-qualified for that major. Is this true? (I haven’t looked.) If the admissions department is, as I suggested earlier, also trying to racially balance specific majors, then that still constitutes discrimination.</p>
<p>I agree that Duke admissions officers’ ratings are more accurate than students’ private judgments. The data seems to further show that Duke is not unduly biased in ratings against Asian applicants (assuming the small discrepancy in personal qualities is insignificant). Does anyone know how Asian applicants actually fare in Duke admissions? Logically the Asian acceptance rate should be marginally higher than the white acceptance rate, perhaps equal once one accounts for legacy and athlete preferences. I address the issue of major “competitiveness” in the previous paragraph, as I understand it.</p>
<p>
</p>
<p>Most of the Asian-heavy STEM majors (science, technology, engineering, mathematics) fields require a higher total SAT at admission. Saying it less politely, white-heavy majors (sociology, art history, etc) require lower total SAT, or only a high verbal SAT, if that. This disparity is true for all sorts of reasons, is well documented, and further evidence that it is enforced at admissions can be seen from the difference in GRE scores of students in different majors. Science vs humanities is not the largest contributor to the white/Asian ratings discrepancy at Duke, but it is one of several factors.</p>
<p>
</p>
<p>That does not follow, no. But if one can find out the Asian share of undergrads at Duke’s engineering school (Pratt), then that would help quantify the expected test-score gap between Asians and whites under a race-neutral model of Duke admissions. The factors other than engineering appear sufficient to account for the data by themselves, but the engineering component is still worth estimating as a way of seeing how much slack there is in reaching that conclusion, or in inferring higher white averages in some categories. </p>
<p>
</p>
<p>No, the Duke data does not show that Asian applicants (or even Asian enrolled students) have personal ratings as high as those of whites. It shows that the white pool is higher in race-neutral preference factors that have race-disparate impact, such as the huge quota of North Carolinians. Unhooked whites may indeed rate higher than unhooked Asians, or it could be a mixture such as higher grades for one and higher test scores for the other.</p>
<p>Rachel Toor, in “Admissions Confidential”, recounting her experience as an admissions officer at Duke, talked about Asian-American applicants looking a lot alike to admissions people, and that this by inference hurt their applications. She also said that admissions people didn’t talk openly about this, but that they tacitly shared the same viewpoint.</p>
<h1>662 to #664, ““And a lot (of Asians) are admitted each year to Princeton, are cross-admits to H, and choose to matriculate to H.””</h1>
<p>Almost all universities have lost <em>some</em> Asian cross-admits to another university. But what you failed to show is that, (a) whether <em>less</em> Asian cross-admits than other ethnic groups, are indeed choosing Princeton over Harvard, and (b) to the extent of lowering the Asian (American) student population at Princeton by more than about 5 %. Furthermore, the departments of economics and sciences are tippy-top at Princeton. Even the department of engineering at Princeton is ranked top 20, about the same as that of Harvard. There is no particular reason why Princeton is less likely favored by Asian cross-admits than by other ethnic groups. The emphasis is on the difference of choices between the ethnics groups, instead of merely Princeton losing some Asian cross-admits to Harvard.</p>
<p>““the Duke data is actually evidence against that school discriminating between white and Asian applicants.””</p>
<p>As a devil’s advocate, if I were part of the-power-to-be and decided to pick on an ethnic group, I would design rules to make discrimination practices look legal and refrain from supplying relevant data, such that the victims are constrained to denial and occasional suspicion. The fact that Duke was willing to release data for publication, a rare occurance among the elites, may suggest that they are probably almost clean, which allows their practices to become a proxy for comparison with other elites (“The Duke Test” as mentioned earlier).</p>
<p>““According to the (Duke) study, the admissions office has a quantifiable form of “private information”, whereas the student’s private information is worthless.””</p>
<p>The is a double-edged sword, because of the lack of transparency. By-and-large as outsiders, we may only infer from observable results. And in a democratic society, the adcoms should reveal more details of the selection process, if only to project the appearance of impartiality. To paraphrase Lord Hewart, “…justice should not only be done, but should manifestly and undoubtedly be seen to be done.”</p>
<p>““No, the Duke data does not show that Asian applicants (or even Asian enrolled students) have personal ratings as high as those of whites. It shows that the white pool is higher in race-neutral preference factors that have race-disparate impact, such as the huge quota of North Carolinians. Unhooked whites may indeed rate higher than unhooked Asians, or it could be a mixture such as higher grades for one and higher test scores for the other.””</p>
<p>A few observations here, (a) “Personal qualities” is inherently a subjective measurement. Only God knows whether it was done in a truly “race-neutral” manner; (b) the discrepancy in the scores (0.05) is too minor to support whether one ethnic group is meaningfully better-rated than the other, especially for a subjective measurement; (c) the Duke data isn’t sufficient to infer a qualified comparison between “unhooked” Asians and “unhooked” Caucasians.</p>
<p>^ The race-neutral preference factors are attributes such as in-state residency, recruited athlete or legacy status, and use of early decision. These have a large effect on the white pool and a small effect on the Asian pool and as such will depress the white ratings in all categories.</p>
<p>
</p>
<p>That not what she wrote. The book is searchable online. Toor did write that:
– Asians with language other than English spoken at home (this was noted for each application) had their verbal scores intepreted more favorably;<br>
– that the admissions staff was ill-suited to process applications from underrepresented minorities (not Asians);<br>
– that immigrant Asians tended to have high test scores and lackluster “personal qualities” ratings, but this was not necessarily true of US Asians;
– the Asian applications from Asian-saturated areas (Northern CA) were highly similar (i.e., the “stereotype” is actually a fact);
– Asians typically had recommendations that complained about their lack of participation in class, that she recognized this as stereotyping and accounted for it in the evaluation;
– Asians often had counselor recommendations that complained about lack of participation in “the life of the school”, but that often those applicants worked in family businesses and the ratings took account of this.</p>
<p>Rachel Toor’s book is a good example of the No Smoking Gun hypothesis. She wrote a hostile kiss-and-tell book in which she trashed Duke, its culture, its admissions office, its faculty quality, pretty much everything. She was an aspiring writer publishing her first book, so the incentive to garner attention was high. Sales would have multiplied, she would have been on a national US talk-show tour and the red carpet treatment all over Asia if she had published evidence of discrimination against Asians. That neither she nor any of the hundreds of people who (used to) work in admissions the past two decades have done so, tells us something. The reports of discrimination all seem to end in the early 1990’s, which is about the same time that statistical signatures of economic discrimination against Asians disappeared.</p>
<p>
</p>
<p>We see the following in the Kim Clark article:</p>
<p>*"Espenshade found that when comparing applicants with similar grades, scores, athletic qualifications, and family history for seven elite private colleges and universities:</p>
<p>Whites were three times as likely to get fat envelopes as Asians".
*</p>
<p>We have already established that Jewish performance on the baseball diamond and in the orchestra are not what set them apart from the Asian applicants. Now we have to examine grades, scores, and family history.</p>
<p>If we are to suggest that Jewish applicants beat Asians in the admissions game because of superior academic qualification, that would suggest gentiles are getting in with relatively low scores because, as suggested by the Duke study, the Asians perform better than whites as a whole. It is possible, but improbable IMO. </p>
<p>If true, I can not see the ACLU and ADL sit idly by. Do you have data to support this?You see why I suspect family history is the reason now?</p>
<p>While still on the Clark article, I also find this quote interesting:</p>
<p>*“We admitted only about half of all the applicants with maximum 2400 SAT scores,” says university spokeswoman Cass Cliatt.
*</p>
<p>I am surprised no one is asking if the other half is composed of Asians. That is why I don’t trust the Adcoms; they are far from honest. </p>
<p>When you add to this the incident of Ms Jones at MIT, and the strange “disappearance” of the guy at Penn, I can see why there is a constant turnover of liberal-minded Adcoms, as suggested by other posters. </p>
<p>
</p>
<p>Considering the source of the suggestion, straighttalk, I am not at all surprised.</p>
<p>““The race-neutral preference factors are attributes such as in-state residency, recruited athlete or legacy status, and use of early decision. These have a large effect on the white pool and a small effect on the Asian pool and as such will depress the white ratings in all categories.””</p>
<p>I would disagree. Caucasians are, afterall, Caucasians, whether they come from in-state, recruited athletes, legacy status, ED, or whatever else. Let me give an analogy, if we were to cherry pick only the Nobel Laureates (race-neutral preference) in the US, then Caucasians would score the highest. Why even bother dissecting the Duke Study. :-)</p>
<p>^re: 678, it makes all the statistical difference if whites are more represented than Asians in these (ostensibly race-neutral) preference categories.</p>
<p>Let’s say that the percentage of white students at Duke who got in under early decision is 30 percent higher than the percentage of Asian students. According to the Early Admissions Game (Avery and coauthors) regressions that’s on average equivalent to a 190 point SAT boost per ED applicant. The impact on the Duke numbers would then be a (0.30)x(190) = 57 SAT point average difference between the Asian and white applicants. The “Duke study” being touted here found only a 47 point difference with some minor differences in other areas, and we haven’t yet taken into account several additional heavy preference categories as well as deficits that specifically affect Asians (clustering by geography and high school, choice of major, etc). </p>
<p>If you run this type of analysis under various assumptions, you wind up with a white/Asian differential on the order of 70 to 100 Espenshade & Chung-style virtual SAT points per applicant. Distributing this in the form of 47 actual SAT points plus marginally higher ratings in several areas, is totally consistent with what the Duke data found. </p>
<p>So what the Duke study shows is something that at first looks like hard data backing the famous Espenshade & Chung “50 point SAT penalty for Asians”, and even a larger penalty considering the other ratings, being explained fairly easily by known race-neutral admissions processes.</p>
<p>
</p>
<p>That’s completely wrong, of course. Espenshade didn’t run any such controlled experiment, and a 3:1 disadvantage for Asians is not what he found. If you believe
he found a 3:1 all-things-equal Asian penalty, you apparently also believe that raising an applicant’s SAT scores by 100 points in some ranges will lower the chances of admission, and that taking more than a couple of AP exams reduces your admissions chances.
Let us know if you agree with such “findings”, which can be found in Espenshade’s logistic regression output linked in post #1 of this thread.</p>